[sc34wg3] CTM: Proposal for %mergemap directive

[Lars Heuer]

Against CTM draft dtd. 2007-11-16 <http://www.jtc1sc34.org/repository/0935.htm>

<http://www.jtc1sc34.org/repository/0935.htm#dir-mergemap>
The %mergemap directive is currently defined as follows:

    mergemap-directive ::= '%mergemap' iri-ref notation?
    notation           ::= iri-ref
    
Where the optional "notation" is again an IRI.


I wonder if it wouldn't make sense to change the syntax to

    notation ::= string

The "string" is case-insensitive and a CTM parser must recognize the
following values: "xtm" and "ctm", but other values like "AsTMa=" etc.
are allowed, too.

Advantages:
- The parser can ignore the EOL delimiter after the mergemap directive
  (this marker is currently necessary to distinguish the notation from
  a subject identifier)
- Syntax is much shorter:
      %mergemap http://example.org/ http://www.topicmaps.org/xtm/
  vs.:
      %mergemap http://example.org/ "xtm"
- The user must not remember a lengthily IRI and potentially typos are
  avoided

Disadvantages:
- An IRI is "more unique" than a string
- A possible version information may be encoded in an IRI more easily


Comments? Thoughts?
  
Best regards,
Lars
-- 
http://www.semagia.com